Why zoologists shouldn’t treat the automatic transmission advantage as an afterthought

Load packages and helper functions

library(tidyverse)
library(kableExtra)
library(DT)

# Create a function to build HTML searchable tables

my_data_table <- function(df){
  datatable(
    df, rownames=FALSE,
    autoHideNavigation = TRUE,
    extensions = c("Scroller",  "Buttons"),
    options = list(
      autoWidth = TRUE,
      dom = 'Bfrtip',
      deferRender=TRUE,
      scrollX=TRUE, scrollY=1000,
      scrollCollapse=TRUE,
      buttons =
        list('pageLength', 'colvis', 'csv', list(
          extend = 'pdf',
          pageSize = 'A4',
          orientation = 'landscape',
          filename = 'inbreeding_studies')),
      pageLength = 100
    )
  )
}

Literature search

To get the gist for the current state of the field, we searched the Web of Science for published journal articles that examined whether animals tolerated, preferred or avoided inbreeding. The following search string was used to search for studies published between the 1st of January 2020 to the 1st of February 2025: TS=(“inbreeding tolerance”) OR TS=(“inbreeding preference”) OR TS=(“inbreeding avoidance”). Query link: https://www.webofscience.com/wos/woscc/summary/e4d61622-9f6b-48f1-b954-83c598431347-014b923814/relevance/1 This search criteria identified 188 studies. We were unable to access two studies and we removed a third from our full-text screen because it was not a journal article.

From this initial list, we identified 121 studies where the authors stated whether they expected inbreeding preference, tolerance or avoidance to be present in their system. Of these, 78 had clear, evolutionary-driven hypotheses regarding inbreeding. These 78 studies were retained for analysis.

To download our reference list you can either download a .csv from the Tables below, get it directly from the source repository or go to our Open Science Network repository.

The dataset

Load and display the complete dataset.

data <- 
  read_delim("inbreeding_adaptive_refs.csv") %>% 
  rename(`Automatic transmission advantage mentioned?` = Automatic_transmission_advantage_mentioned,
         Title = Article_Title,
         Journal = Source_Title,
         Year = Publication_Year,
         Relevance = Relevant,
         Type = Article_Type,
         `Publication format` = `Publication Type`,
         `Null hypothesis` = Null_expectation)

my_data_table(data %>% select(Authors, Title, Journal, Year, `Null hypothesis`, `Automatic transmission advantage mentioned?`, Taxa, Type, Relevance, Taxa, Accessed, `Publication format`))

Column explanations

• Authors: authors listed on publication

• Title: publication title

• Journal: journal where study is published

• Year: year study was published

• Null hypothesis: did the authors expect inbreeding preference, tolerance or avoidance to be present in their system. ‘No evolutionary statement’ indicates that they did not posit an evolutionary hypothesis regarding the presence or absence of inbreeding.

• Automatic transmission advantage mentioned: did the authors, whether by name or description, mention the automatic transmission advantage when making their evolutionary argument about inbreeding.

• Taxa: animal or plant.

• Type: was the study Empirical, a review, a meta-analysis or theoretical?

• Relevance: how central was inbreeding to the studies research question. 0 = not relevant, 1 = not central or 2 = central, or they clearly stated a evolutionary hypothesis about inbreeding in their system.

• Accessed: 1 = we found the full text, 0 = we did not.

• Publication format: J = journal article, B = book chapter.

Clean the dataset and display data used for analysis

analysis_data <- 
  data %>% 
  filter(Accessed == 1, # remove studies we couldn't access (also includes the book chapter)
         Relevance == 2, # scored 0, 1, 2, only include 2s
         `Null hypothesis` != "No evolutionary statement", # remove studies that mention inbreeding, but make no clear evolutionary prediction relating to it
        Taxa == "Animal") %>%
  select(Authors, Title, Journal, Year, `Null hypothesis`, `Automatic transmission advantage mentioned?`, Taxa, Type, Relevance, Taxa, Accessed, `Publication format`)

Show the studies selected for analysis

my_data_table(analysis_data)

Visualise the data

analysis_data %>%
  group_by(`Automatic transmission advantage mentioned?`, 
           `Null hypothesis`) %>% 
  summarise(Studies = n()) %>% 
  knitr::kable(digits = 1, align = "lccc", position = "c") %>%
  kable_styling("striped") 

Automatic transmission advantage mentioned?	Null hypothesis	Studies
NO	Inbreeding avoidance	53
NO	Inbreeding preference/tolerance	1
NO	None	6
YES	Inbreeding avoidance	9
YES	Inbreeding preference/tolerance	3
YES	None	6

Test whether mentioning the ATA changes the ‘gist’

Using a Fisher’s exact test, we test if there is a relationship between mentioning the ATA and whether inbreeding avoidance is the null expectation.

analysis_data_for_test <-
  analysis_data %>% 
  mutate(Hypothesis = case_when(
    `Null hypothesis` == "Inbreeding avoidance" ~ 1,
    `Null hypothesis` != "Inbreeding avoidance" ~ 0)) %>% 
  rename(ATA = `Automatic transmission advantage mentioned?`)

test <- fisher.test(table(analysis_data_for_test$ATA, analysis_data_for_test$Hypothesis))

test

    Fisher's Exact Test for Count Data

data:  table(analysis_data_for_test$ATA, analysis_data_for_test$Hypothesis)
p-value = 0.00123
alternative hypothesis: true odds ratio is not equal to 1
95 percent confidence interval:
 0.03309415 0.52695725
sample estimates:
odds ratio 
 0.1369227